Vespper

⭐ Star us on Github & follow us on Twitter & LinkedIn!

📜 TL;DR

Vespper is an on-call engineer that helps engineers troubleshoot alerts by surfacing the right data at the right time.

https://youtu.be/uN888sOaIig

🎯 What is Vespper?

Vespper is an on-call engineer running 24/7 to troubleshoot your alerts and surface the right data to help you resolve your issue so incidents never falls through the cracks.

Whether your company is going through growing pains, needs a better handle at dealing with SEV0 or has too many low priority unsolved issues - Vespper will adapt to your needs to democratize expert knowledge across your organization.

❌ The problem

Most companies drown in alerts and there are too many alerts to handle.

• Engineers hate waking up at 3am to go investigate an issue, one person is familiar with one specific thing and they aren’t around, there is a constant need to check for false positives.
• Product managers are left playing hot potato when trying to figure out the impact of incidents and bugs but engineers are too busy triaging the issue itself.
• Customer support are left asking “who owns” which service when triaging tickets.
• Engineering leaders need to constantly ensure engineers are escalating quickly, they rely on service specialists to debug an issue other engineers aren’t familiar with but they aren’t always around during critical times.

💰 The solution

Vespper is a multi-agent system that triages alerts, troubleshoots them and sends findings in seconds to Slack. It’s connected to internal tools (observability, incident management, knowledge management, codebases and more) and can surface problems & identify patterns in the oceans of data you have.

At the moment, we support popular tools such as:

• Observability - DataDog, Grafana, Prometheus, Jaeger, SigNoz, Coralogix
• Incident management - PagerDuty, Opsgenie
• Codebases - Github
• Knowledge management - Notion, Confluence, Jira
• Communication -  Slack, Teams (TBD)

⚒️ How does it work?

Vespper is a system that’s comprised of multi-agents and AIOps models. Behind the scenes, we run & coordinate multiple agents and tools that help identify suspicious patterns from your environment.

Using Vespper is easy.

1. Signup & create an organisation.
2. Connect the tools you work with when triaging/troubleshooting. This means observability, knowledge, incident management and communication tools.

Once the integrations are connected, we automatically trigger advanced data ingestion pipelines that starts scraping data from your environment. This is used to train the bot. You can see the status of the ingestion & training in our web UI.

Once the system is ready, it will start triaging alerts for you, post an hypothesis on Slack and show you all the automatic checks it made!

👨‍👨‍👦 About the team

Hi, we are Topaz & Dudu 😊 We co-founded Vespper with the mission of making on-call stress-free.

Topaz (CEO) - Spent years at Snyk (a hyper-growth unicorn) building and maintaining large distributed systems, leading PLG experiments and achieving 99.9% uptime for her teams services.

Dudu (CTO) - has seven years of experience working at rapidly-growing technology startups, including Google, Viz.ai and SafeBreach. Dudu was a deep-learning algorithm engineer at Viz.ai. He contributed to cutting-edge projects that leverage computer vision for healthcare applications and also worked on large-scale distributed systems.

We both dealt with daily alerts and the daily monotony of triaging, tuning and maintaining services observability.  This time spent all came at the expense of working on more impactful customer focused work.

We believe that pairing the strength of AI to pattern-match, classic AIOps and heuristics from our own experience can unlock a new experience for developers.

🙏 Our ask

1. Book a meeting with us https://cal.com/callwithtopaz/vespper or email us at founders@vespper.com.
2. Star us on GitHub + follow along on Twitter & LinkedIn.